Analysis of human and mouse reprogramming of somatic cells to induced pluripotent stem cells. What is in the plate?

Abstract

After the hope and controversy brought by embryonic stem cells two decades ago for regenerative medicine, a new turn has been taken in pluripotent cells research when, in 2006, Yamanaka's group reported the reprogramming of fibroblasts to pluripotent cells with the transfection of only four transcription factors. Since then many researchers have managed to reprogram somatic cells from diverse origins into pluripotent cells, though the cellular and genetic consequences of reprogramming remain largely unknown. Furthermore, it is still unclear whether induced pluripotent stem cells (iPSCs) are truly functionally equivalent to embryonic stem cells (ESCs) and if they demonstrate the same differentiation potential as ESCs. There are a large number of reprogramming experiments published so far encompassing genome-wide transcriptional profiling of the cells of origin, the iPSCs and ESCs, which are used as standards of pluripotent cells and allow us to provide here an in-depth analysis of transcriptional profiles of human and mouse cells before and after reprogramming. When compared to ESCs, iPSCs, as expected, share a common pluripotency/self-renewal network. Perhaps more importantly, they also show differences in the expression of some genes. We concentrated our efforts on the study of bivalent domain-containing genes (in ESCs) which are not expressed in ESCs, as they are supposedly important for differentiation and should possess a poised status in pluripotent cells, i.e. be ready to but not yet be expressed. We studied each iPSC line separately to estimate the quality of the reprogramming and saw a correlation of the lowest number of such genes expressed in each respective iPSC line with the stringency of the pluripotency test achieved by the line. We propose that the study of expression of bivalent domain-containing genes, which are normally silenced in ESCs, gives a valuable indication of the quality of the iPSC line, and could be used to select the best iPSC lines out of a large number of lines generated in each reprogramming experiment.

Figure 1. Timeline of publications in the reprogramming field.

Timeline of publication of reprogramming papers in mouse and human, with a simplified classification of the main message/achievement of each paper. See supplementary File S1 for a more detailed and updated description of published reprogramming reports.

Figure 2. Human protein-protein interaction networks of genes with higher expression levels in ESCs and iPSCs compared to somatic cells.

The human protein-protein interaction networks of genes most consistently highly expressed in ESCs and iPSCs, compared to the starting cell populations, have been created from the lists of the biggest changes in expression, using String with high confidence interactions (min score 0.7) and have been edited in Medusa. They show a central, highly interconnected network of genes in which the most famous pluripotency transcription factors are to be found and which is likely to represent the core pluripotency network. They also highlight a number of genes whose functions relate to cell-cell communication, cell cycle, DNA repair and other metabolisms.

Figure 3. Mouse protein-protein interaction networks of genes with higher expression levels in ESCs and iPSCs compared to somatic cells.

The mouse protein-protein interaction networks of genes most consistently highly expressed in ES and iPSCs, compared to the starting cell populations, have been created from the lists of biggest changes in expression, using String with high confidence interactions (min score 0.7) and have been edited in Medusa. They show a central, highly interconnected network of genes in which the most famous pluripotency transcription factors are to be found and which is likely to represent the core pluripotency network. They also highlight a number of genes whose functions relate to cell-cell communication, cell cycle, DNA repair and other metabolisms.

Figure 4. Number of genes which may be problematic for further differentiation of mouse iPSC lines generated by different laboratories.

Number of bivalent domain-containing genes for each iPS cell line which show some expression in the iPS cell whereas they are silent in 100% or at least 80% of available ESC lines analyzed, and therefore could influence the differentiation potential of the iPS cell lines.

Figure 5. Barriers to reprogramming.

The process of somatic cell reprogramming entails overcoming the cellular barriers that preserve cell identity. The first barrier consists of the stress generated by the overexpression of factors that stimulates apoptosis and reduces cell viability. The p53 pathway is an important factor for this barrier. Many cells that overcome this barrier end up trapped in a partially reprogrammed state in which they are able to self-renew but are not yet pluripotent, as reflected by their ability to form tumors when injected into immunosuppressed mice. These cells are dependent on the presence of the transfactors and cannot activate the expression of the endogenous pluripotency factors due to the presence of a non-permissive chromatin environment on their regulatory regions, constituting a second barrier to reprogramming. Only after overcoming this barrier are cells fully pluripotent and able to produce teratomas after injection into immunodepressed mice.